Abstract:The metal-insulator transition has been a subject of intense research since Nevil Mott has first proposed that the metallic behavior of interacting electrons could turn to the insulating one as electron correlations increase. Here, we consider electrons with massless Dirac-like dispersion in two spatial dimensions, described by the Hubbard models on two geometrically different lattices, and perform numerically exact calculations on unprecedentedly large systems that, combined with a careful finite size scaling… Show more
“…By extrapolating the magnetization to the thermodynamical limit, Ref. [1] located the critical point at U/t = 3.78, a value in line with recent numerical simulations [75,76]. According to the present study, the pinning field is a relevant perturbation for the line critical behavior, so that for h 0 = 0, under the RG the model flows away from the "ordinary" fixed point h 0 = 0.…”
Section: Discussion and Future Directionssupporting
confidence: 87%
“…By means of MC simulations, we have checked this scenario for both the bilayer Heisenberg model and the improved Blume-Capel model, and verified the exact result for the scaling dimension of the pinning field at the h 0 = 0 fixed point. This picture is also expected to hold for the Hubbard model on the honeycomb lattice, which undergoes a quantum phase transition in the Gross-Neveu Heisenberg UC [74][75][76]. For this model the scaling dimension of the pinning field is found to be significantly smaller than for the O(N) models, such that very large lattice sizes would be needed in order to reach the asymptotic behavior; see the discussion in Appendix B.…”
Section: Discussion and Future Directionsmentioning
We apply a recently advocated simulation scheme that employs a local order-parameter pinning field to study quantum critical phenomena in the two-dimensional square-lattice bilayer quantum Heisenberg model. Using a world-line quantum Monte Carlo approach, we show that for this model, the pinning-field approach allows to locate the quantum critical point over a wide range of pinning-field strengths. However, the identification of the quantum critical scaling behavior is found to be hard since the pinning field introduces strong corrections to scaling. In order to further elucidate the scaling behavior in this situation, we also study an improved classical lattice model in the three-dimensional Ising universality class by means of Monte Carlo simulations on large lattice sizes, which allow us to employ refined finite-size scaling considerations. A renormalization group analysis exhibits the presence of an important crossover effect from the zero pinning-field to a critical adsorption fixed point. In line with field-theoretical results, we find that at the critical adsorption fixed point the shortdistance expansion of the order-parameter profile exhibits a new universal critical exponent. This result also implies the presence of slowly decaying scaling corrections, which we analyze in detail.
“…By extrapolating the magnetization to the thermodynamical limit, Ref. [1] located the critical point at U/t = 3.78, a value in line with recent numerical simulations [75,76]. According to the present study, the pinning field is a relevant perturbation for the line critical behavior, so that for h 0 = 0, under the RG the model flows away from the "ordinary" fixed point h 0 = 0.…”
Section: Discussion and Future Directionssupporting
confidence: 87%
“…By means of MC simulations, we have checked this scenario for both the bilayer Heisenberg model and the improved Blume-Capel model, and verified the exact result for the scaling dimension of the pinning field at the h 0 = 0 fixed point. This picture is also expected to hold for the Hubbard model on the honeycomb lattice, which undergoes a quantum phase transition in the Gross-Neveu Heisenberg UC [74][75][76]. For this model the scaling dimension of the pinning field is found to be significantly smaller than for the O(N) models, such that very large lattice sizes would be needed in order to reach the asymptotic behavior; see the discussion in Appendix B.…”
Section: Discussion and Future Directionsmentioning
We apply a recently advocated simulation scheme that employs a local order-parameter pinning field to study quantum critical phenomena in the two-dimensional square-lattice bilayer quantum Heisenberg model. Using a world-line quantum Monte Carlo approach, we show that for this model, the pinning-field approach allows to locate the quantum critical point over a wide range of pinning-field strengths. However, the identification of the quantum critical scaling behavior is found to be hard since the pinning field introduces strong corrections to scaling. In order to further elucidate the scaling behavior in this situation, we also study an improved classical lattice model in the three-dimensional Ising universality class by means of Monte Carlo simulations on large lattice sizes, which allow us to employ refined finite-size scaling considerations. A renormalization group analysis exhibits the presence of an important crossover effect from the zero pinning-field to a critical adsorption fixed point. In line with field-theoretical results, we find that at the critical adsorption fixed point the shortdistance expansion of the order-parameter profile exhibits a new universal critical exponent. This result also implies the presence of slowly decaying scaling corrections, which we analyze in detail.
“…Past numerical simulations of graphene suggest that there is a phase transition between semi-metal and Mott insulator that occurs near U c /κ ≈ 3.5 [16]. More recent calculations have pushed this value larger, U c /κ ≈ 3.8 [12]. Since matching the nanotube data favors a lower U, our results suggest that graphene is a semi-metal.…”
Section: Comparison With Data and Discussionsupporting
confidence: 63%
“…Our correlation functions are accurate to O(δ 2 ) [15], and so the calculated effective masses should scale as O(δ), which motivates the use of Eq. (12). We combine our extrapolated points in the right panel of Figure 4 and show the dependence of ∆/κ as a function of Hubbard ratio U/κ for different tube geometries.…”
Abstract. We show how lattice Quantum Monte Carlo simulations can be used to calculate electronic properties of carbon nanotubes in the presence of strong electron-electron correlations. We employ the path integral formalism and use methods developed within the lattice QCD community for our numerical work and compare our results to empirical data of the Anti-Ferromagnetic Mott Insulating gap in large diameter tubes.
“…DQMC [15][16][17] has been widely used in the investigation of correlated fermion systems [18][19][20][21][22][23][24][25][26][27][28][29][30][31][32]. Despite of the great successes, the method also suffers from serious difficulties.…”
Self-learning Monte Carlo method is a powerful general-purpose numerical method recently introduced to simulate many-body systems. In this work, we extend it to interacting fermion quantum system in the framework of widely used determinantal quantum Monte Carlo. The new method can generally reduce the computational complexity, and moreover can greatly suppress the autocorrelation time near a critical point. This enables us to simulate interacting fermion system on a 100 × 100 lattice even at the critical point for the first time, and obtain critical exponents with high precision.Introduction -Numerous intermetallic compounds hosting intriguing phenomena such as non-Fermi liquid[1] and unconventional superconductivity [2-4] emerging from quantum critical fluctuations (antiferromagnetic [2-4], nematic [5,6], etc), demand proper theoretical understanding of them. However, these systems are usually strongly correlated, and can only be solved by non-perturbative methods. In the last few years, after several attempts [7][8][9][10][11][12][13][14], people realize determinantal quantum Monte Carlo (DQMC) is one of the most suitable methods and sometimes even the only available choice.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.